Optimized Backstepping Tracking Control Using Reinforcement Learning for Quadrotor Unmanned Aerial Vehicle System

Wen, Guoxing; Hao, Wei; Feng, Wei; Gao, Kaizhou

doi:10.1109/tsmc.2021.3112688

Cited by 47 publications

(16 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Driven by rapid developments of emerging technologies such as artificial intelligence, extended reality, and blockchain, metaverse is becoming an attainable reality. As a promising artificial intelligence technique, deep reinforcement learning (DRL) recently achieves remarkable success in both video games of virtuality [2], [3] and many real-world scenes, such as robotic manipulation [4], [5], mobile robot control [6], [7], [8], [9], [10], [11], [12], [13], and manufacturing process [14], [15], which makes it ideally suited for the realization of metaverse intelligence. Multiagent DRL (MARL) is a multiagent extension of DRL that concentrates on the relation and interaction of multiple agents in mixed cooperativecompetitive environments [16].…”

Section: Introductionmentioning

confidence: 99%

MARL Sim2real Transfer: Merging Physical Reality With Digital Virtuality in Metaverse

Shi

Liu

Zhang

et al. 2023

IEEE Trans. Syst. Man Cybern, Syst.

View full text Add to dashboard Cite

Metaverse is an artificial virtual world mapped from and interacting with the real world. In metaverse, digital entities coexist with their physical counterparts. Powered by deep learning, metaverse is inevitably becoming more intelligent in the interactions between reality and virtuality. However, it is confronted with a nontrivial problem known as sim2real transfer when deep learning techniques try to bridge the reality gap between the physical world and simulations. In this article, we use multiagent deep reinforcement learning (MARL) to implement collective intelligence for digital entities as well as their physical counterparts. To model the immersive environments in metaverse, we define a nonstationary variant of Markov games and propose a recurrent MARL solution to it. Based on the solution, MARL sim2real transfer that bridges real and virtual multiple unmanned aerial vehicle (multi-UAV) systems is successfully conducted by employing recurrent multiagent deep deterministic policy gradient (R-MADDPG) with the domain randomization technique. Additionally, we use perception-control modularization to improve the generalization performance of MARL policies and make training more efficient.

show abstract

Section: Introductionmentioning

confidence: 99%

MARL Sim2real Transfer: Merging Physical Reality With Digital Virtuality in Metaverse

Shi

Liu

Zhang

et al. 2023

IEEE Trans. Syst. Man Cybern, Syst.

View full text Add to dashboard Cite

show abstract

“…When considering the presence of dead zones and disturbances in the actuators, the difficulty of the prescribed-time position control problem of AGVs will increase significantly [4,27]. In the future, leader-follower consensus control [28] or tracking control [29] of vehicle formations using reinforcement learning is also a worthwhile research topic.…”

Section: Discussionmentioning

confidence: 99%

Prescribed‐time stabilisation control of differential driven automated guided vehicle

Chen

Zhang

2022

IET Cyber-Syst and Robotics

View full text Add to dashboard Cite

The position control problem of differential‐driven automated guided vehicles (AGVs) based on the prescribed‐time control method is studied. First, an innovative orientation error function is proposed by an auxiliary arcsine function about the orientation angle. Thus, the problem of position control of AGV is transformed into the stabilisation control of the kinematic system. Second, by introducing a reserved time parameter and a smooth switching function, a novel time‐varying scaling function is proposed. This novel scaling function avoids the risk of infinite gain in the conventional prescribed‐time control method while ensuring the smoothness of control laws. Then, an improved velocity constraint function is proposed using the Gaussian function. Compared with the existing constraint function, the proposed method not only has more smoothness but also solves the balance point errors caused by the large AGV orientation errors. The presented method ensures that the AGV reaches the target position in a prescribed time. Hence, the upper bound of the AGV system state can be determined by adjusting parameters. Matlab simulation results show that the proposed controller can effectively make the AGV system state satisfy the prescribed bound.

show abstract

“…In an effort to overcome the limitations of linear control methods, nonlinear control techniques have been widely shown to achieve an increased level of reliability in quadrotor flight under various uncertain and adversarial operating conditions. Common nonlinear control design approaches include feedback linearization [8], [9], backstepping [10], [11], [12], adaptive control [13], [14] and robust control [15], [16], [17]. Advantages to feedback linearization include mathematical simplicity and easy implementation; however, they often incur reduced robustness to disturbances and sensitivity to model uncertainty.…”

Section: Introductionmentioning

confidence: 99%

Adaptive Modified RISE-based Quadrotor Trajectory Tracking with Actuator Uncertainty Compensation

Kidambi¹,

Tiwari²,

Ijoga³

et al. 2023

Preprint

View full text Add to dashboard Cite

This paper presents an adaptive robust nonlinear control method, which achieves reliable trajectory tracking control for a quadrotor unmanned aerial vehicle in the presence of gyroscopic effects, rotor dynamics, and external disturbances. Through novel mathematical manipulation in the error system development, the quadrotor dynamics are expressed in a control-oriented form, which explicitly incorporates the uncertainty in the gyroscopic term and control actuation term. An adaptive robust nonlinear control law is then designed to stabilize both the position and attitude loops of the quadrotor system. A rigorous Lyapunov-based analysis is utilized to prove asymptotic trajectory tracking, where the region of convergence can be made arbitrarily large through judicious control gain selection. Moreover, the stability analysis formally addresses gyroscopic effects and actuator uncertainty. To illustrate the performance of the control law, comparative numerical simulation results are provided, which demonstrate the improved closed-loop performance achieved under varying levels of parametric uncertainty and disturbance magnitudes.

show abstract

Optimized Backstepping Tracking Control Using Reinforcement Learning for Quadrotor Unmanned Aerial Vehicle System

Cited by 47 publications

References 36 publications

MARL Sim2real Transfer: Merging Physical Reality With Digital Virtuality in Metaverse

MARL Sim2real Transfer: Merging Physical Reality With Digital Virtuality in Metaverse

Prescribed‐time stabilisation control of differential driven automated guided vehicle

Adaptive Modified RISE-based Quadrotor Trajectory Tracking with Actuator Uncertainty Compensation

Contact Info

Product

Resources

About